Wrap allocated GPU memory in numpy-compatible arrays.
authorTomasz Rybak <tomasz.rybak@gmail.com>
Mon, 21 Mar 2011 14:00:14 +0000 (15:00 +0100)
committerMichał Januszewski <spock@gentoo.org>
Thu, 24 Mar 2011 20:55:54 +0000 (21:55 +0100)
commitf6bd3207593cdc2009dc8f7acd254ab874417f50
tree820ede7a188693d7be8d3fcc5e800ae76e8ab132
parent16929f1b7c738f7c6299db631199344c9d83a815
Wrap allocated GPU memory in numpy-compatible arrays.

Add ability to run reduction kernels on GPU.
sailfish/backend_cuda.py
sailfish/backend_dummy.py
sailfish/backend_opencl.py
sailfish/lb_binary.py
sailfish/lbm.py